1. Estimation Questions How do you estimate? What are you going to estimate? Where do you start?

2. Estimation Techniques - Traditional Project Management Approaches  Guesstimating  Delphi Technique  Time Boxing  Top-Down  Bottom-Up  Analogous Estimates (Past experiences)  Parametric Modeling (Statistical)

3. Guestimating Estimation by guessing or just picking numbers out of the air is not the best way to derive a project’s schedule and budget. Unfortunately, many inexperienced project managers tend to guesstimate, or guess at the estimates, because it is quick and easy.

4. Delphi Technique  Involves multiple, anonymous experts  Each expert makes an estimate  Estimates compared  If close, can be averaged  If not, do another iteration until consensus is reached

5. Time Boxing  A “box” of time is allocated for a specific activity, task, or deliverable  Can focus a team if used effectively  Can demoralize a team if not used effectively

6. Top-Down  Top & middle managers determine overall project schedule &/or cost  Lower level managers are expected to breakdown schedule/budget estimates into specific activities (WBS)

7. Bottom-Up  Schedules & budgets are constructed from WBS  Starts with people who will be doing the work  Schedules & budgets are the aggregate of detailed activities & costs

8. Analogous Estimates  Similar to Top-Down approach  Use information from previous, similar projects as a basis for estimation

9. Parametric Modeling  Use project characteristics (parameters) in a mathematical model to estimate  Example: $50/ LOC based on:  Programming language  Level of expertise  Size & complexity

10. 6.2 Test Results Report 6.2.1 Review test plan with client1 day 6.2.2 Carry out test plan5 days 6.2.3 Analyze results2 days 6.2.4 Prepare test results report and presentation3 days 6.2.5 Present test results to client1 day 6.2.6 Address any software issues or problems5 days Estimates are made for each activity in the WBS How did we come up with these estimates? Using a technique, or combination of techniques, with the exception of guestimating!

11. Estimating Techniques - Software Engineering Approaches  Lines of Code (LOC)  Function Points  COCOMO  Heuristics Software engineering techniques focus on estimating the size of the system to be developed

12. Determinants of Estimating the Largest Deliverable of the Project – The Application System

13. Function Point Analysis  Allan Albrecht, IBM – 1979  Synthetic metric  Independent of the Technology  IFPUG standards (www.ifpug.org)  5 Primary Elements  Inputs  Outputs  Inquiries  Logical Files  Interfaces

14. The Application Boundary for Function Point Analysis

15. Complexity LowAverageHighTotal Internal Logical Files (ILF) _3 x 7 = 21_2 x 10 = 20_1 x 15 = 1556 External Interface Files (EIF) __ x 5 = ___2 x 7 = 14__ x 10 = __14 External Input (EI) _3 x 3 = 9_5 x 4 = 20_4 x 6 = 2453 External Output (EO) _4 x 4 = 16_2 x 5 = 10_1 x 7 = 733 External Inquiry (EQ) _2 x 3 = 6_5 x 4 = 20_3 x 6 = 1844 Total Unadjusted Function Points (UAF) 200

16. General System CharacteristicDegree of Influence Data Communications3 Distributed Data Processing2 Performance4 Heavily Used Configuration3 Transaction Rate3 On-line Data Entry4 End User Efficiency4 Online Update3 Complex Processing3 Reusability2 Installation Ease3 Operational Ease3 Multiple Sites1 Facilitate Change2 Total Degrees of Influence 40 Value Adjustment Factor VAF = (TDI * 0.01) +.65 VAF = (40 *.01) +.65 = 1.05 Total Adjusted Function Points = FP = UAF * VAFFP = 200 * 1.05 = 210

17. LanguageAverage Source LOC per Function Pont Average Source LOC for a 210 FP Application Access387,980 Basic10722,470 C12826,880 C++5311,130 COBOL10722,470 Delphi296,090 Java5311,130 Machine Language640134,440 Visual Basic 5296,090 Source: http://www.spr.com/library/0langtbl.htm

18. COCOMO  COnstructive COst MOdel  Developed by Barry Boehm,  Has been extended to COCOMO II  http://sunset.usc.edu/csse/research/COCOMOII/cocomo _main.html http://sunset.usc.edu/csse/research/COCOMOII/cocomo _main.html

19. COCOMO Models (Effort)  Organic – Routine  Person Months = 2.4 * KDSI 1.05  Embedded – Challenging  Person Months = 3.6 * KDSI 1.20  Semi-Detached – Middle  Person Months = 3.0 * KDSI 1.12

20. COCOMO – Effort Example  Semi-Detached 10,600 Java LOC = 200 FP * 53 Person Months = 3.0 * KDSI 1.12 = 3.0 * (10.6) 1.12 = 42.21

21. COCOMO Models (Duration)  Organic  Duration = 2.5 * Effort 0.38  Semi-Detached  Duration = 2.5 * Effort 0.35  Embedded  Duration = 2.5 * Effort 0.32

22. COCOMO Duration Example Duration = 2.5 * Effort 0.35 = 2.5 *(42.21) 0.35 = 9.26 months People Required = Effort / Duration = 42.21 / 9.26 = 4.55

23. Heuristics (Rules of Thumb) When scheduling a software task: 1/3 – Planning 1/6 – Coding 1/4 – Component test and early system test 1/4 – System test, all components in hand

24. The Man Month Months People Months People Time versus number of workers perfectly partitionable task – i.e., No communication among them e.g., reaping wheat. When a task that cannot be partitioned because of sequential constraints, the application of more effort has no effect on the schedule.

25. Adding People  Increases the total effort necessary  The work & disruption of repartitioning  Training new people  Added intercommunication

26. What can cause inaccurate estimates?  Scope changes  Overlooked tasks  Poor developer-user communication  Poor understanding of project goals  Insufficient analysis  No (or poor) methodology  Changes in team  Red tape  Lack of project control  Not identifying or understanding impact of risks

27. Other Factors to Consider When Estimating  Rate at which requirements may change  Experience & capabilities of project team  Process or methods used in development  Specific activities to be performed  Programming languages or development tools to be used  Probable number of bugs or defects & removal methods  Environment or ergonomics of work space  Geographic separation of team across locations  Schedule pressure placed on the team

28. How can estimates be improved?  Experience!  Lessons learned  Best Practices  Revision  Monitor  Focus on deliverables  Control